I was at a conference in Jan 2023.  Dr Josh McGrath was a speaker. He was employed by OCP North America; however, I think he was previously a soil extension specialist at U of Kentucky.  He talked about soil sampling. One of the things I recall he said was that they had done research (at U of Kentucky, I believe) where they did just that - pull soil samples, and then moved over a couple of inches and pulled another set of soil samples, and compared the results.  They had the same results - there was variation in the results even when location was moved a couple of inches.  He said pull 8 cores and calling it a sample was not sufficient. He recommended 18-20 cores (I believe) to get a good representative sample.  Definitely more cores is better, to get around this variation we get in the soils.  That is one "nugget" I recall from his presentation.

If you get a chance to hear Dr. McGrath give a presentation on soil sampling, I recommend you take advantage of it. He is worth hearing on the subject.